This is a past event
Earth's climate can be viewed as a large dynamical system evolving in time.
Earth's climate can be viewed as a large dynamical system evolving in time. Modelling often gives a set of deterministic differential equations (possibly with light noise), akin to those of a dissipative mechanical system. So the types of (codimension–one) bifurcation that can be generically encountered under the slow variation of a single control are well understood and classified. This could help in the study of climate tipping points that are of current concern: and we note that most of these bifurcations display the useful precursor that the decay rate of a critical de–stabilizing mode vanishes at the bifurcation.
One aim of global warming research is to estimate the proximity of the climate, or one of its sub-systems, to such a tipping point. Key sub–systems have been described by Lenton (2008) who calls them tipping elements. One identified tipping element, the North Atlantic thermohaline circulation (THC) exhibits a fold bifurcation. Under the slow increase of fresh water flux (from rivers and melting ice) into the North Atlantic a jump from this fold can trigger a suddenly shut down of the system.
In this talk, we present work by Held & Kleinen on this shut–down. These authors use the oceanic output of a coupled climate model of intermediate complexity. They make a 50,000 yrs run with a linear increase in atmospheric CO2 which generates within the model an increase in the fresh water forcing. The run ends with the total collapse of the THC. Although this could not be predicted by watching for the folding of the 'equilibrium path' they show that the time-series of a mapping coefficient, c, allows a fairly good prediction of the imminent collapse using a linear extrapolation. Indeed the steady rise of c towards its critical value of +1 at the fold is detectable over a very considerable time scale
- Speaker
- Professor J Michael T Thompson FRS
- Venue
- FN, G011